run-41/checkpoint-1290/trainer_state.json

{
  "best_metric": 0.8978286157425359,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-41/checkpoint-1290",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1290,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.1879228353500366,
      "learning_rate": 0.0001500473390748166,
      "loss": 0.31,
      "step": 645
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8937572702597906,
      "eval_f1": 0.1189710610932476,
      "eval_loss": 0.288364440202713,
      "eval_mcc": 0.19117679563293483,
      "eval_precision": 0.6981132075471698,
      "eval_recall": 0.06502636203866433,
      "eval_runtime": 9.9344,
      "eval_samples_per_second": 519.205,
      "eval_steps_per_second": 16.307,
      "step": 645
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5299654006958008,
      "learning_rate": 0.00012861200492127138,
      "loss": 0.2852,
      "step": 1290
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8978286157425359,
      "eval_f1": 0.20030349013657056,
      "eval_loss": 0.2860374450683594,
      "eval_mcc": 0.2650150871029742,
      "eval_precision": 0.7333333333333333,
      "eval_recall": 0.11599297012302284,
      "eval_runtime": 9.9216,
      "eval_samples_per_second": 519.878,
      "eval_steps_per_second": 16.328,
      "step": 1290
    }
  ],
  "logging_steps": 500,
  "max_steps": 5160,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "total_flos": 3173885065200.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.886564686264361,
    "learning_rate": 0.00017148267322836185,
    "num_train_epochs": 8,
    "temperature": 15
  }
}